A Mechanism for Value-Sensitive Decision-Making

Abstract

We present a dynamical systems analysis of a decision-making mechanism inspired by collective choice in house-hunting honeybee swarms, revealing the crucial role of cross-inhibitory ‘stop-signalling’ in improving the decision-making capabilities. We show that strength of cross-inhibition is a decision-parameter influencing how decisions depend both on the difference in value and on the mean value of the alternatives; this is in contrast to many previous mechanistic models of decision-making, which are typically sensitive to decision accuracy rather than the value of the option chosen. The strength of cross-inhibition determines when deadlock over similarly valued alternatives is maintained or broken, as a function of the mean value; thus, changes in cross-inhibition strength allow adaptive time-dependent decision-making strategies. Cross-inhibition also tunes the minimum difference between alternatives required for reliable discrimination, in a manner similar to Weber's law of just-noticeable difference. Finally, cross-inhibition tunes the speed-accuracy trade-off realised when differences in the values of the alternatives are sufficiently large to matter. We propose that the model, and the significant role of the values of the alternatives, may describe other decision-making systems, including intracellular regulatory circuits, and simple neural circuits, and may provide guidance in the design of decision-making algorithms for artificial systems, particularly those functioning without centralised control.

Towards a Cognitive Design Pattern for Collective Decision-Making

“…On the one hand, this may lead to errors in the decision making if the system happens to be in the basin of attraction of the inferior choice. On the other hand, as also noted in [11], larger cross-inhibition rates lead to increased decision speed (see Figure 5(a)). Similar dynamics can be observed for different values of θ < π, for which we observed that smaller differences between the alternatives sometimes lead to the wrong choice, as predicted by the macroscopic model (see supplementary material http://iridia.ulb.ac.be/supp/IridiaSupp2014-005/sm.pdf).…”

“…It is worth noting that this model does not require any explicit comparison of the alternatives' quality by the single individuals. The quality value of the two alternativeshereafter labelled v A and v B -is instead encoded in the transition rates (e.g., through value-dependent discovery or recruitment rates [11]): different-quality alternatives correspond to biased transition rates, while same-quality alternatives to unbiased ones. Overall, the collective decision is based purely on the system dynamics resulting from individual-to-individual interactions.…”

“…An extensive analysis of this model showed that the cross-inhibition rates σ i crucially determine the dynamics of decision-making (see [11]). In case of same-quality alternatives (v A = v B ), the transition rates for different alternatives have the same value (i.e., γ i = γ, ρ i = ρ, α i = α and σ i = σ).…”

“…Additionally, in case of options of different quality, the cross-inhibition rate determines the minimum quality difference between the two alternatives to break the symmetry and make a systematic choice [11]. Therefore, there exist parameterisations of the system that allow to obtain accurate decisions when the quality of the options differs sufficiently, or random decisions when the values are similar.…”

“…This process requires the estimation of the quality of the available alternatives (possibly with uncertainty) to select the best one. Recent studies have identified optimal decision strategies in decentralised systems [8,13,11].We propose a cognitive design pattern hinged on these studies, to be applied to artificial distributed systems. Although decision-making has been largely studied in this context [2,7,9,12], general purpose solutions are still missing.…”